Messages in this thread | | | Date | Fri, 04 Feb 2000 21:01:15 +0100 | From | Manfred Spraul <> | Subject | Re: 2.3.42 alpha updates |
| |
Richard Henderson wrote: > > On Fri, Feb 04, 2000 at 08:42:15PM +0100, Manfred Spraul wrote: > > It seems that the Alpha port still calls schedule() during tlb flush > > operations. Could you please remove the pointer_lock() from > > smp_call_function()? Just replace it with a simple spin_lock() or > > spin_lock_bh(). > > Actually, just replace the schedule with a barrier. Using a > spinlock would require extra storage, which would be wasteful. >
So you saved 4 bytes. What about saving 20+ lines and killing pointer_lock?
SMP development is difficult enough, and I prefer if everyone uses a few primitives (semphore, spinlock, rw_sem, rw_lock) instead of special functions.
And if you use spin_lock_bh(), then we would allow smp_call_function() from bottom half handlers - might be important for hardware drivers. i386 allows that. -- Manfred
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |